arXiv:2511.09170v1 Announce Type: new 
Abstract: This article presents HOTFLoc++, an end-to-end framework for LiDAR place recognition, re-ranking, and 6-DoF metric localisation in forests. Leveraging an octree-based transformer, our approach extracts hierarchical local descriptors at multiple granularities to increase robustness to clutter, self-similarity, and viewpoint changes in challenging scenarios, including ground-to-ground and ground-to-aerial in forest and urban environments. We propose a learnable multi-scale geometric verification module to reduce re-ranking failures in the presence of degraded single-scale correspondences. Our coarse-to-fine registration approach achieves comparable or lower localisation errors to baselines, with runtime improvements of two orders of magnitude over RANSAC for dense point clouds. Experimental results on public datasets show the superiority of our approach compared to state-of-the-art methods, achieving an average Recall@1 of 90.7% on CS-Wild-Places: an improvement of 29.6 percentage points over baselines, while maintaining high performance on single-source benchmarks with an average Recall@1 of 91.7% and 96.0% on Wild-Places and MulRan, respectively. Our method achieves under 2 m and 5 degrees error for 97.2% of 6-DoF registration attempts, with our multi-scale re-ranking module reducing localisation errors by ~2$\times$ on average. The code will be available upon acceptance.

يقدم المقال HOTFLoc++، وهو إطار عمل جديد للتعرف على الأماكن باستخدام LiDAR وتحديد المواقع المترية 6-DoF في الغابات. يعزز الإطار من المتانة ضد الفوضى وتغيرات الزاوية، محققًا معدل استرجاع 90.7% على CS-Wild-Places، متفوقًا بشكل كبير على الطرق الحالية. هذه التطورات مهمة لتطبيقات الملاحة الذاتية ومراقبة البيئة، حيث تحسن من دقة وكفاءة تحديد المواقع في البيئات المعقدة.

El artículo presenta HOTFLoc++, un nuevo marco para el reconocimiento de lugares LiDAR y la localización métrica 6-DoF en bosques. Mejora la robustez frente al desorden y los cambios de perspectiva, logrando un Recall@1 del 90.7% en CS-Wild-Places, superando significativamente los métodos existentes. Este avance es crucial para aplicaciones de navegación autónoma y monitoreo ambiental, ya que mejora la precisión y eficiencia de la localización en entornos complejos.

L'article présente HOTFLoc++, un nouveau cadre pour la reconnaissance de lieux LiDAR et la localisation métrique 6-DoF dans les forêts. Il améliore la robustesse face au désordre et aux changements de point de vue, atteignant un Recall@1 de 90,7 % sur CS-Wild-Places, surpassant considérablement les méthodes existantes. Cette avancée est cruciale pour les applications de navigation autonome et de surveillance environnementale, car elle améliore la précision et l'efficacité de la localisation dans des environnements complexes.

The article introduces HOTFLoc++, a novel framework for LiDAR place recognition and 6-DoF metric localization in forests. It enhances robustness against clutter and viewpoint changes, achieving a Recall@1 of 90.7% on CS-Wild-Places, significantly outperforming existing methods. This advancement is crucial for applications in autonomous navigation and environmental monitoring, as it improves localization accuracy and efficiency in complex environments.

HOTFLoc++: End-to-End Hierarchical LiDAR Place Recognition, Re-Ranking, and 6-DoF Metric Localisation in Forests

Was this article worth reading? Share it

LucidQuery AI

Attentive AI

Lenso.ai

Deptho.ai

Octofy

Com.locatelloapp

Ready to build your own newsroom?